Nina van Leeuwen (30) throws an auger over her shoulder and walks onto the forest path. Her zip-off pants, raincoat and wellies are covered in dried mud. For her PhD research at the Vrije Universiteit Amsterdam (VU), she is investigating old river courses of the Overijsselse Vecht to map prehistoric floods. She has been living in a mobile home at a campsite near Ommen for two weeks, so she can go into the field six days a week.
Published at the end of September an investigation from the VU Geophysical Research Letterswhich investigated the same thing but near the Maas. Willem Toonen (41) was co-author of this and he now supervises Van Leeuwen in her research. He is also present today. Just like Charlotte Bischot (22), master’s student earth sciences at the VU, who helps with the fieldwork. Ultimately, Toonen and Van Leeuwen want to combine the research, and Toonen’s previous work on the Rhine, to see whether the same mechanisms cause extreme flooding in the different rivers. Not only to know what the rivers did in the past, but especially what they can do in the future.
Because only in the last seventy years or so have there been measurements of flooding in the Overijsselse Vecht, says Toonen. Measurements of the Rhine and Maas started around 1900. As a result, there is great uncertainty in flood predictions. “But there are thousands of years of data underground.” This allows rare floods to be better mapped. Take the major floods of the Maas in 2021, says Toonen. Through data from historical sources and the subsurface, he knows that an equally large flood took place in 1643. “But it is not yet included in the high water statistics.” Research into the Meuse produced more than 3,500 years of data: from 1500 BC to the 19th century, with evidence for more such major floods.
After a hundred meters, Toonen suddenly stops. He points to a clearing, to the right of the path, half a meter lower. “This is an old bend, the river once ran here. It flooded two years ago.” The current river is a few hundred meters further, where the brown field ends and a green dike runs.
The research into the Meuse showed that flooding of the river is related to a meteorological phenomenon, the North Atlantic Oscillation (NAO). “We are now in the middle of one stormtrack” says Toonen, pointing to the heavily cloudy gray sky. It is a few days after Storm Benjamin and one low pressure area after another is sweeping across the country. How storms approach Europe is determined by the air pressure difference between Iceland and the Azores, he explains: the NAO. If that difference is large – low air pressure above Iceland and high pressure above the Azores – then the westerly winds are strong and storms are coming our way in quick succession.
Willem Toonen and Nina van Leeuwen conduct research along the Vecht.
Photo Wouter de Wilde
Grass, earth, peat and gyttja
Climate change may change the strength of the NAO, says Toonen: if the ocean warms, the difference in temperature and air pressure between the tropics and the poles may become less significant. The westerly winds and the jet stream higher in the atmosphere, which drives the low-pressure areas, will then become less strong, causing storms to move more slowly and therefore hang over the same place for longer. Showers can also become more intense in a warmer climate. This could lead to more local precipitation and larger floods, such as in the summer of 2021.
At the next clearing, Van Leeuwen scans the barbed wire fence. On an accessible patch of grass, she places the auger on the bottom wire and climbs through. The rest follows. We are standing in an old meander of the Overijsselse Vecht, which is now about two hundred meters away. Nowadays, due to human intervention, the river looks more like a canal, but when a river is left to take its course, it meanders through the landscape. Because “there is no such thing as justice,” says Toonen. And with the slightest bit of curvature, the flow speed on the outside is higher than on the inside. “And then he will bite more out of the bank there.” This results in longer and longer bends, until the river cuts itself off again.
The grass is soggy, there are large puddles from the heavy rain of the past few days. This did not matter for the fieldwork, says Van Leeuwen. “I just had to take my notes like this,” she says, shielding an imaginary notebook from the rain with her upper body. She sticks the end of a yellow measuring tape into the ground with a kind of rounded knife. She rolls it out a few meters towards the river. She then routinely turns the auger into the grass with Bischot – Van Leeuwen has already done around 150 drillings in recent weeks, between Ommen and Hardenberg. She makes notes of the layers of sediment she finds for each drilling and takes samples with her to the lab.
When the drill comes out of the ground, Van Leeuwen places the gouge, a thin tube, next to the measuring tape. She scrapes off the mud, revealing the layers beneath. She smooths the layers with her finger. The first piece is grass and soil, she points out, then comes peat. Then gyttja“a Swedish word for organic sediment deposited in a standing water environment,” from the time when the old meander was again cut off from the river and there was a small lake at this spot. Every now and then there is a layer of sand in between: the floods. When the river overflowed its banks, it took sand with it and deposited it in the lake. Van Leeuwen: “The larger the grain of sand, the more current and the greater the flood.”
Nina van Leeuwen (right) and Charlotte Bischot drill for remains of past floods.
Photo Wouter de Wilde
Hanging years on layers
After each drilling, Van Leeuwen attaches a new steel tube to the drill so that it can go one meter deeper. With every additional meter she drills, the drill swings more dangerously back and forth in the air as she pulls it out. Van Leeuwen previously looked at the old channels with ground-penetrating radar to see their width and depth. This way she can estimate the amount of water that once flowed through the channel. And to know how deep the drilling should be: here she drills four meters deep, then she hits the old river bottom.
She points to a dark spot with thin layers in the sediment. “Here you can see the flat plant remains.” She holds up a black flake on her fingertip: a piece of ancient leaf. The veins are still clearly visible. She puts it in a plastic jar to take with her to the lab. That kind of organic material is essential for her research: thanks to the leaf remains, she can hang years on the sediment layers.
When a leaf dies, the absorption of carbon from the air stops. Carbon with isotope 14 (not normal carbon-12, which makes up 99 percent of all carbon atoms) begins to decay radioactively with a known half-life: 5,736 years. If you dig it up again centuries later and run it through a mass spectrometer, you can calculate back how old the plant remains are – accurately within about forty years. That method, carbon dating, only works for about 40,000 years, Toonen adds; after that is too much 14C expired. But that is more than enough for this study, which is limited to the Holocene; the geological epoch that begins 11,700 years ago, after the last ice age.
They cannot directly date the sand layers from the floods. But the layers in between are, and this is how the earth scientists can make a timeline of the floods. The relative size of the floods can be read from the coarseness of the sediment. Van Leeuwen characterizes them relative to each other and calculates the average number of years in which they occur.
I think it’s very special that you can learn so much about what a landscape was like in the past
Van Leeuwen’s research is mainly aimed at linking prehistoric floods to the climate, and comparing them with the Maas and Rhine to see whether the patterns in floods occur over a large area. In the past it has also been relatively warm a number of times, says Toonen. In the late Middle Ages, for example, and in Roman times. Periods for which Rijkswaterstaat has no data, but the subsurface does. Toonen: “We may see floods that we have not seen before, people are now saying. But that is a big assumption: we do not really know what we have seen in the past.”
Although practical applications for the Maas and Vecht are still far away, the same type of research has already made a difference in the Rhine. Toonen: “The data for the Rhine have not changed the insights very much. We did not think: oh, one meter can be removed from the dikes, or one meter should be added.” But the uncertainty has been reduced in the flood forecasts. Dikes are designed for the most extreme high water levels. Based on the known measurements, the uncertainty regarding the chance of an extreme flood was quite large, because the measurement series was relatively short. By adding historical floods, researchers from the University of Twente still arrived at approximately the same amount of extreme discharge, but with less uncertainty, says Toonen.
“We do have contact with Rijkswaterstaat and Deltares and we share our information. But this is independent fundamental research: to better understand the systems and to demonstrate that we can do this.” It can be applied to rivers worldwide, Toonen continues. “It is not yet happening on a large scale, but this is innovative.”
“I think it is very special that through these types of sediment layers you can learn so much about what a landscape was like in the past,” says Van Leeuwen. “And that by comparing it with other things you can even find out what the influence of climate change was.” Floods in particular are an important topic, she believes. “Researchers generally agree that flooding will become more extreme in the future. I think it’s cool to contribute, even if only in a small way, to understanding that better.”
The Vecht near Ommen.
Photo Wouter de Wilde
NEW: Give this item as a gift
As an NRC subscriber you can subscribe every month 10 items give as a gift to someone without an NRC subscription. The recipient can read the article directly, without a paywall.